FR3107411A1 - Integrated circuit comprising an adaptation and filtering network and corresponding adaptation and filtering method. - Google Patents

Abstract

The integrated circuit comprises a power amplifier (PA) intended to supply a signal in a fundamental frequency band, an antenna (ANT), and a matching and filtering network (MFN) comprising: - a first section (SCT1) , a second section (SCT2), and a third section (SCT3); the three sections having LC assemblies configured to present an impedance matched to the output of the power amplifier (PA) in the fundamental frequency band, in which the LC assemblies of the first section (SCT1) and the second section ( SCT2) are further configured to have resonant frequencies respectively adapted to attenuate the harmonic frequency bands of the fundamental frequency band. Figure for the abstract: Fig 1

Description

Integrated circuit comprising an adaptation and filtering network and corresponding adaptation and filtering method.

Embodiments and implementations relate to integrated circuits comprising a matching and filtering network, typically between an output of a power amplifier and an antenna.

On the one hand, power amplifiers operate at maximum efficiency when the load on their outputs is at optimum impedance. The optimum output impedance of a power amplifier typically differs from the impedance of the antennas connected to it.

Therefore, impedance matching circuits are conventionally provided between the output of power amplifiers and the antennas in order to transform the impedances of the antennas to the ideal impedances of the amplifiers. Impedance matching circuits are typically realized by passive components such as resistive elements, inductive elements, and capacitive elements. Impedance matching circuits consume a potentially bulky surface area, which can be particularly troublesome when the impedance matching circuit is implemented in a small integrated circuit.

On the other hand, the emission spectra of power amplifiers typically contain spurious signals, such as harmonic frequencies of the fundamental frequency or noise due to the internal non-linearity of power amplifiers.

Consequently, filter circuits are conventionally provided between the output of the power amplifiers and the antennas in order to filter in particular the harmonic frequency bands, for example up to the fifth order. The filter circuits are typically produced by passive components such as resistive elements, inductive elements, and capacitive elements, and themselves also consume a surface which can be bulky, and particularly troublesome when the impedance matching circuit is carried out. in an integrated way.

Impedance matching and filtering circuits have specific constraints and are conventionally designed separately.

Although the techniques separating the two functions (impedance matching and filtering) are satisfactory from a performance standpoint, these techniques require many passive components, which makes them expensive and cumbersome.

However, the reduction in the size of the integrated circuits reduces the surface available for the passive devices, which makes it more difficult to integrate the two functions with satisfactory performance.

The use of surface wave filters "SAW" (for "Surface Acoustic Wave" in English), reduces the space but increases the cost.

Therefore, there is a need to provide compact and inexpensive solutions for implementing impedance matching and filtering with good performance.

According to embodiments and embodiments, it is proposed to ensure the impedance matching and the filtering in a single network, in which the impedance matching is ensured by a network with a minimum quality factor, in order to obtain the best response over the widest possible frequency range while keeping the number of passive elements as low as possible; filtering is provided by replacing passive elements with inductive-capacitive assemblies judiciously chosen to resonate in the frequencies to be filtered without modifying the transformation of impedance in the emission frequencies. The resonance frequencies of the inductive-capacitive assemblies are in particular chosen so as to optimize the size and the response of the network.

In this regard, there is proposed according to one aspect an integrated circuit comprising a power amplifier intended to supply a signal in a fundamental frequency band, an antenna, and a matching and filtering network comprising:
- a first section between an output node of the power amplifier and a first intermediate node,
- a second section between the first intermediate node and a second intermediate node, and
- a third section between the second intermediate node and an input node of the antenna,
the three sections comprising inductive-capacitive "LC" assemblies configured to present an impedance adapted to the output of the power amplifier in the fundamental frequency band. The LC assemblies of the first section and of the second section are further configured to have resonant frequencies respectively adapted to attenuate the harmonic frequency bands of the fundamental frequency band.

For convenience and according to conventional usage in electronics, the notation "LC" will be used to denote the terms "inductive-capacitive".

Thus, there is proposed a three section matching and filtering network of LC assemblies, in which all the LC assemblies of the first section and of the second section are configured both to match the impedance and to filter the frequencies. harmonics.

It will be noted in particular that all the LC assemblies of each of the first and second sections are configured for the simultaneous adaptation and filtering functions, unlike the conventional techniques in which one assembly is dedicated to the adaptation and another assembly, at least. partially separate, is dedicated to filtering.

The third section is dedicated to impedance matching, and in particular is not configured to have a resonant frequency, in order to maintain attenuation in the high frequencies.

Consequently, the design according to this aspect proposing a complete fusion of the adaptation and filtering functions in the LC assemblies of the first section and of the second section allows a particularly compact realization of the integrated circuit, without losing performance or increasing costs. .

According to one embodiment, the LC assemblies of the first section are configured to have resonance frequencies lower than the resonance frequencies of the corresponding LC assemblies of the second section.

This is because for impedance matching, the capacitive elements are usually larger on the power amplifier side in the first section than on the antenna side in the second section. On the other hand, the resonant frequency of an LC arrangement is inversely proportional to the size of the inductive and capacitive elements of the LC arrangement. Consequently, this embodiment proposes to position the resonance frequencies in an optimized manner for the sizes of the inductive elements provided for the filtering. The overall size is thus optimized as low as possible.

According to one embodiment, the first section comprises a parallel LC circuit coupled between the output node of the power amplifier and the first intermediate node and a series LC circuit coupled between the first intermediate node and a ground node, and the second section has a parallel LC circuit coupled between the first intermediate node and the second intermediate node and a series LC circuit coupled between the second intermediate node and the ground node.

Thus, parallel LC arrangements block signals at their resonant frequencies, while series LC arrangements ground signals at their resonant frequencies.

According to one embodiment, each LC assembly in series is configured to have an equivalent impedance corresponding to an impedance of a capacitive element suitable for said impedance matching in the fundamental frequency band, and in which the resonant frequencies of each LC series connections are chosen so as to be distributed in different harmonic frequency bands of the fundamental frequency band, and so that the series LC connections having equivalent impedances corresponding to the impedances of the capacitive elements having the smallest values capacitive, have the highest resonance frequencies.

In other words, the resonant frequency of the LC assemblies in series in the first section and in the second section is chosen so as to introduce an inductive element of minimum size in combination with a capacitive element provided for impedance matching. .

Indeed, given that the value of the inductive element necessary to make a capacitive element resonate is inversely proportional to the capacitive value and to the square of the resonant frequency, the largest resonant frequencies are advantageously associated with the smallest capacitive values. , to minimize the value of the inductive element to add.

This embodiment again proposes to optimize the overall size by positioning the resonance frequencies in an optimized manner for the sizes of the inductive elements allowing the filtering function in each of the first and second sections.

According to one embodiment, the resonant frequencies of each LC assembly in parallel are chosen so as to be distributed, with the resonance frequencies of the LC assemblies in series, in different harmonic frequency bands of the fundamental frequency band.

Thus, this embodiment makes it possible to cover filtering on all harmonic frequency bands, by positioning the resonance frequencies in the harmonic frequency bands remaining to be filtered, by combining capacitive elements with inductive elements provided. for impedance matching.

Indeed, given that the inductive elements are typically much more bulky than the capacitive elements, the positioning of the resonant frequencies with the capacitive value in the LC assemblies in parallel, introduces an additional bulk of acceptable magnitude compared to the bulk. equivalent inductive elements provided for impedance matching, and also with respect to the space saving by optimizing the inductive elements in the LC filters in series.

According to one embodiment, the LC assemblies of the first section and of the second section are configured according to at least one of the following criteria:
the parallel LC assembly of the first section is configured to have a resonant frequency in one half of the frequency band of the second harmonics;
- the LC assembly in parallel of the second section is configured to have a resonant frequency in the other half of the frequency band of the second harmonics;
- The series LC assembly of the first section is configured to have a resonant frequency in the third harmonic frequency band;
- the LC circuit in series of the second section is configured to have a resonant frequency either between the frequency band of the fourth harmonics and the frequency band of the fifth harmonics, or in a common portion of the frequency band of the fourth harmonics and of the frequency band of fifth harmonics.

This embodiment provides possibilities for positioning the resonant frequencies to optimize the overall size, and for optimal performance.

In particular, the positioning of the resonant frequency of the series LC assembly of the first section in the third harmonic frequency band, and not in the upper half of the second harmonic frequency band, will be noted. This positioning advantageously makes it possible to avoid possible problems of coupling between the resonant frequency of the LC assembly in parallel with the first section (which is in the frequency band of the second harmonics), with the resonant frequency of the LC assembly in series. of the first section (which is not in the frequency band of the second harmonics).

According to one embodiment, the third section comprises an LC assembly comprising an inductive element coupled between the second intermediate node and the input node of the antenna, and a capacitive element coupled between the input node of the antenna and a ground node, the LC assembly of the third section being configured to have a minimum quality factor.

According to another aspect, there is provided a method of impedance matching and filtering between an output of a power amplifier supplying a signal in a fundamental frequency band and an antenna, comprising a dimensioning of a network of virtual adaptation comprising:
- a first section between an output node of the power amplifier and a first intermediate node,
- a second section between the first intermediate node and a second intermediate node, and
- a third section between the second intermediate node and an input node of the antenna,
the three sections each comprising an inductive element and a capacitive element, the dimensioning being carried out so as to present an impedance suitable for the output of the power amplifier in the fundamental frequency band,
in which the method comprises a realization of a real matching and filtering network comprising a replacement of each inductive element and of each capacitive element of the first section and of the second section of the virtual matching network, by inductive arrangements -respective "LC" capacitors, configured to present an equivalent impedance matched to the output of the power amplifier in the fundamental frequency band, and further to have resonant frequencies respectively matched to attenuate the harmonic frequency bands of the fundamental frequency band.

The method according to this aspect provides a dimensioning of a virtual impedance matching network, not providing the filtering function, in order to dimension the requirements for the impedance matching.

By virtual is meant for example "which exists in the state of mere possibility or eventuality, but not in the state of material realization", as in particular as an intermediary of calculations.

The filtering is then introduced into the real adaptation and filtering network, that is to say, as opposed to "virtual", which actually exists for the implementation of the adaptation and the filtering on a signal d 'emission. This real matching and filtering network is obtained by replacing the virtual elements with equivalent real elements for the purposes of impedance matching, and having in addition their filtering functions.

This process makes it possible to achieve in practice, for any realization of the power amplifier and the antenna, the complete fusion of the adaptation and filtering functions in the LC assemblies of the first section and of the second section in a compact manner. , efficient and inexpensive.

According to one embodiment, the resonance frequencies of the LC assemblies of the first section of the real matching and filtering network are chosen lower than the resonance frequencies of the corresponding LC assemblies of the second section of the matching network and of actual filtering.

According to one embodiment, the first section of the virtual adaptation network comprises an inductive element coupled between the output node of the power amplifier and the first intermediate node and a capacitive element coupled between the first intermediate node and a ground node, and the second section of the virtual adaptation network comprises an inductive element coupled between the first intermediate node and the second intermediate node and a capacitive element coupled between the second intermediate node and the ground node,
and wherein said realization of the actual matching and filtering network comprises replacing each inductive element with an LC connection in parallel, and replacing each capacitive element with an LC connection in series.

According to one embodiment, the resonant frequencies are first chosen for each LC assembly in series so as to be distributed in different harmonic frequency bands of the fundamental frequency band, and so that the LC assemblies in series replacing the capacitive elements of the virtual matching network having the smallest capacitive values, have the largest resonance frequencies.

According to one embodiment, the resonant frequencies are then chosen for each LC assembly in parallel so as to be distributed, with the resonance frequencies of the LC assemblies in series, in different harmonic frequency bands of the band. fundamental frequencies.

According to one embodiment, the realization of the real adaptation and filtering network is made according to at least one of the following criteria:
- the LC assembly in parallel with the first section has a resonant frequency in one half of the frequency band of the second harmonics;
- the parallel LC assembly of the second section has a resonant frequency in the other half of the frequency band of the second harmonics;
- the LC assembly in series of the first section has a resonant frequency in the frequency band of the third harmonics;
- the LC assembly in series of the second section has a resonant frequency either between the frequency band of the fourth harmonics and the frequency band of the fifth harmonics, or in a common portion of the frequency band of the fourth harmonics and of the band frequencies of fifth harmonics.

According to one embodiment, the realization of the real adaptation and filtering network comprises a reproduction of the third section of the virtual adaptation network comprising an inductive element coupled between the second intermediate node and the input node of the antenna, and a capacitive element coupled between the input node of the antenna and a ground node, the LC assembly of the third section being dimensioned to have a minimum quality factor.

Other advantages and characteristics of the invention will become apparent on examination of the detailed description of embodiments and implementation, which are in no way limiting, and of the accompanying drawings in which:

illustrate embodiments and implementation of the invention.

The illustrates an MFN adaptation and filtering network between an output node of a power amplifier PA and an input node of an antenna ANT, for example produced in an integrated manner in an integrated circuit.

The power amplifier PA is configured to provide a transmission signal in a fundamental frequency band, in particular radio frequencies suitable for wireless communications, such as, for example, telecommunications of the 4G, 5G or LTE type, Wifi or still bluetooth.

The MFN matching and filtering network has three matching and filtering sections SCT1, SCT2, SCT3, and a DCFD direct current supply stage.

The DCFD direct current supply stage has an inductive element in series between a supply voltage terminal VCC and the output node of the power amplifier, and a capacitive element between the supply voltage terminal VCC and a ground reference voltage terminal GND.

The DCFD direct current power stage is used to supply the voltage and current level required for the MFN matching and filtering network from the output node of the PA power amplifier.

The first section SCT1 is located between the output node of the power amplifier PA and a first intermediate node N1, the second section SCT2 is located between the first intermediate node N1 and a second intermediate node N2, and the third section SCT3 is located between the second intermediate node N2 and an input node of the ANT antenna.

Each of the three sections SCT1, SCT2, SCT3 comprises inductive elements and capacitive elements, that is to say inductive-capacitive assemblies which will be designated for convenience and conventionally "LC" assemblies.

The LC assemblies of the three sections SCT1, SCT2, SCT3 are configured to present an impedance suitable for the output of the power amplifier PA in the fundamental frequency band.

The impedance is matched to the output of the power amplifier PA in that at this impedance, the optimum power of the transmission signal is transferred from the power amplifier PA to the antenna ANT, in particular in order to ensure minimum reflection of the transmit signal power.

In addition, the LC assemblies of the first section SCT1 and the second section SCT2 are configured to have resonant frequencies respectively adapted to attenuate the harmonic frequency bands of the fundamental frequency band.

Harmonic frequencies are integer multiples of the fundamental frequency of the send signal.

In particular, the first section SCT1 comprises a parallel LC assembly 11 coupled between the output node of the power amplifier PA and the first intermediate node N1, as well as a series LC assembly 12 coupled between the first intermediate node N1 and a GND ground node.

Similarly, the second section SCT2 comprises a parallel LC circuit 21 coupled between the first intermediate node N1 and the second intermediate node N2, as well as a series LC circuit 22 coupled between the second intermediate node N2 and the ground node. GND.

The third section SCT3 comprises an inductive element coupled between the second intermediate node N2 and the input node of the ANT antenna, as well as a capacitive element coupled between the input node of the ANT antenna and the GND ground.

The LC assembly of the third section is configured to have a minimum quality factor, that is to say that it is not intended to provide a filtering function on a resonant frequency, but nevertheless makes it possible to maintain a attenuation in high and very high frequencies. The LC setup in the third section is dedicated to impedance matching.

It will be noted that a DC coupling capacitor is provided in a conventional manner between the input node of the antenna ANT and the antenna, in order to block the DC component of the voltage, and its capacitive value is chosen sufficiently large to have negligible impact on impedance matching.

Thus, the parallel LC assemblies 11, 21, will block the transmission of the signals at their resonance frequencies along the series path going from the output of the power amplifier PA to the antenna ANT, by means of the intermediate nodes N1, N2.

And, the series LC assemblies 12, 22, will evacuate to the GND ground (usually "shunt" in English) the signals at their resonance frequencies flowing from the output of the power amplifier PA to the antenna ANT, by the intermediary of the intermediate nodes N1, N2.

According to an advantageous exemplary embodiment, each LC assembly in parallel 11, 21 is configured in a dual manner so as on the one hand to have an equivalent impedance corresponding to an impedance of an inductive element intended for said impedance adaptation in the frequency band. fundamental ( ), and on the other hand to have a resonant frequency f11, f21 ( ) chosen in one of the harmonic frequency bands ( ).

Similarly, each LC assembly in series 12, 22 is advantageously configured in a dual manner so as, on the one hand, to have an equivalent impedance corresponding to an impedance of a capacitive element suitable for said impedance adaptation in the fundamental frequency band ( ), and to have a resonant frequency f12, f22 ( ) chosen in one of the harmonic frequency bands ( ) to minimize the inductive value of the LC assembly in series at this equivalent impedance.

It will be noted in particular that each LC assembly 11, 12, 21, 22 of the first section SCT1 and of the second section SCT2 are in their entirety configured simultaneously for the adaptation and filtering functions. That is to say that there is no component in the LC assemblies of the first section SCT1 and of the second section SCT2 which are only dedicated to the impedance matching function or that to the filter function.

Reference is made to FIGS. 2 to 4 to present an advantageous method of dimensioning the inductive elements and the capacitive elements of the impedance matching and MFN filtering network described above in relation to the .

The illustrates a virtual adaptation network MN0 which will serve as a reference base for dimensioning the inductive and capacitive elements of the adaptation and filtering network MFN described in relation to the .

The MN0 matching network is qualified as “virtual” because this network has only a computational vocation to size the needs for impedance matching. The sizing results will serve as a basis for calculation to evaluate the components actually produced to implement the adaptation and filtering by the MFN network described in relation to the .

The MN0 virtual adaptation network has three virtual adaptation sections SCT01, SCT02, SCT03, and a DCFD direct current power supply stage.

The three virtual sections SCT01, SCT02, SCT03 are similar to the structure of the third section SCT3, i.e. they each have an inductive element on the serial channel going from the output node of the power amplifier PA up to the input node of the ANT antenna, via the intermediate nodes N1, N2, as well as a capacitive element coupled to the ground GND, in "shunt", and on the intermediate nodes N1, N2 and the node ANT antenna input.

The virtual adaptation network MN0 thus corresponds to a low-pass network with a minimum quality factor in three sections SCT01, SCT02, SCT03, designed to adapt the impedance between the output of the power amplifier PA and the antenna ANT .

We refer to the .

The represents a Smith chart normalized by the impedance of the ANT antenna, so that the impedance of the RANT antenna is located in the center of the Smith chart.

The sizing is done in such a way as to transform the impedance of the RANT antenna to the ideal impedance of the RPA power amplifier.

For this, intermediate impedances R1, R2 are calculated by geometric mean between the impedance of the antenna RANT and the ideal impedance RPA presented on the output node of the power amplifier PA.

That is to say:
R1 = (RA² x RL) ^1/3
R2 = (RA x RL²) ^1/3

Strictly speaking, the calculation is made with RPA being the inverse of the real part of the admittance presented at the output node of the power amplifier PA, and RANT being the inverse of the real part of the admittance presented by the ANT antenna.

The intermediate impedances R1, R2 correspond to the impedances (strictly speaking the reverse of the real parts of the admittances) which will be presented on the intermediate nodes N1, N2 of the virtual adaptation network MN0.

The values of capacitive elements C ₀ and inductive elements L ₀ are derived for each section SCT01, SCT02, SCT03 (or SCTk - ) by reading the Smith chart, and by the equations EQ1 and EQ2 defined in relation to the .
EQ1:
EQ2:
With ω the pulsation at a frequency (f ₀ ) chosen in the fundamental frequency band, R _L the impedance (strictly speaking the inverse of the real part of the admittance) presented to the left of each section SCTk (k∈ [01; 02; 03]) and R _G the impedance (strictly speaking the reverse of the real part of the admittance) presented to the right of each SCTk section, as illustrated by .

The imaginary part of the admittance of the load on the output node of the power amplifier PA is made by the inductive element of the DC power supply stage DCFD, by which the power amplifier PA is powered.

The method then comprises a realization of the real adaptation and filtering network MFN, as described previously in relation with the , from the inductive L0 and capacitive C0 elements thus dimensioned in each section SCT01, SCT02, SCT03 of the virtual adaptation network MN0.

Reference is made in this regard to Figures 1 and 2.

The realization of the real adaptation and filtering network MFN comprises a replacement of each inductive element and of each capacitive element of the first section SCT01 and of the second section SCT02 of the virtual adaptation network MN0, by inductive-capacitive assemblies " LC ”respective resonants. The inductive-capacitive "LC" arrangements are configured to present an impedance equivalent to the matched impedance of the virtual matching network MN0 in the fundamental frequency band, and further to have resonant frequencies respectively matched to attenuate the bands of harmonic frequencies of the fundamental frequency band.

In particular, the inductive elements L ₀ of the virtual adaptation network MN0 are replaced by parallel LC assemblies 11, 12, having the same impedance at a frequency f ₀ chosen within the fundamental frequency band, and a respective resonant frequency f _r.

The inductive values L and the capacitive values C of the components of the parallel LC assemblies 11, 21 are given, for each section SCT1, SCT2, by the equations EQ3 and EQ4.
EQ3:
EQ4:
With L ₀ the inductive value of the inductive element respectively replaced in each section SCT01, SCT02, f ₀ a frequency in the fundamental frequency band, and f _r the resonant frequency of the respective parallel LC circuit.

And in particular, the capacitive elements C ₀ of the virtual matching network MN0 are replaced by series LC assemblies 12, 22, having the same impedance at a frequency f ₀ chosen within the fundamental frequency band, and a resonance at a respective _{resonant frequency f r.}

The capacitive values C and the inductive values L of the components of the LC assemblies in series 12, 22 are given, for each section SCT1, SCT2, by the equations EQ5 and EQ6
EQ5:
EQ6:
With C ₀ the capacitive value of the capacitive element respectively replaced in each section SCT01, SCT02, f ₀ a frequency in the fundamental frequency band, and f _r the resonant frequency of the respective parallel LC circuit.

The resonance frequencies f _r are chosen so as to attenuate the harmonic frequency bands of the fundamental frequency band.

We refer to the .

The represents results of the MFN adaptation and filtering network as described in relation to the and obtained as described in relation to FIGS. 2 to 4, with an advantageous positioning of the resonant frequencies (fr) f11, f12, f21, f22.

Graph 51 represents the gain in transmission of the adaptation and filtering network MFN, graph 52 represents the gain in transmission of the MFN network in the fundamental frequency band FB, graph 53 represents the real part of the impedance of the network MFN in the fundamental frequency band FB, and graph 54 represents the imaginary part of the impedance of the MFN network in the fundamental frequency band FB.

The resonant frequency of the LC assembly in parallel 11 of the first section SCT1 is denoted f11, the resonant frequency of the LC assembly in series 12 of the first section SCT1 is denoted f12, the resonant frequency of the LC assembly in parallel 21 of the second section SCT2 is denoted f21, the resonant frequency of the LC assembly in series 22 of the second section SCT2 is denoted f22 (with reference to ). The different resonance frequencies will be designated directly by their respective references.

In an advantageous example, the resonant frequency f11 is positioned in one half, for example the lower half, of the frequency band of the second harmonics HB2.
The resonant frequency f12 is positioned in the frequency band of the third harmonics HB3.
The resonant frequency f21 is positioned in the other half, for example the upper half, of the frequency band of the second harmonics HB2.
The resonant frequency f22 is positioned between the frequency band of the fourth harmonics HB4 and the frequency band of the fifth harmonics HB5, that is to say, in the event of overlap of the frequency bands of the fourth harmonics HB4 and the fifth harmonics HB5, in the common portion of said HB4 and HB5 bands.

This allows, on the one hand, to provide more than 38dB of attenuation to the frequency band of the second harmonics HB2 and more than 40 dB to all other harmonics.

In addition, this example corresponds to an optimization in terms of space requirements for the realization of the adaptation and MFN filtering network.

Indeed, given that the inductive elements have a size and a much larger footprint than the capacitive elements in this type of embodiment; that the capacitive elements C ₀ coupled to ground in the virtual matching network MN0 have lower values near the antenna ANT and higher near the power amplifier PA; and that the value of the inductive element required to make a capacitive element resonate is inversely proportional to the capacitive value and to the square of the resonant frequency; it is the inductive elements added to make the capacitive elements coupled to the ground resonate that should be minimized.

The smallest capacitive value C _{0 is} therefore first associated with the largest resonant frequency, in this example between the frequency bands of the fourth harmonics HB4 and of the fifth harmonics HB5. This makes it possible to minimize the value of the inductive element L, determined by the equation EQ6, of the LC assembly in series 22 of the second section SCT2.

Then, we choose to make the capacitive element of the first section SCT01 of the virtual network MN0 resonate in the highest harmonic frequency band which is not yet filtered, that is to say the frequency band of third harmonics HB3. This makes it possible to minimize the value of the inductive element L, determined by the equation EQ6, of the series LC assembly 12 of the first section SCT1, while additionally covering the frequency band of the third harmonics HB3.

Thus, the set of two inductive elements of the series LC assemblies 12, 22 of the two sections SCT1, SCT2 has a minimal overall size, while having resonance frequencies respectively distributed in a suitable manner to attenuate several bands of harmonic frequencies. HB3, HB4, HB5.

The resonance frequencies of the parallel LC 11 and T 30 assemblies can then be positioned more freely in the frequency bands of the remaining harmonics HB2, since the additional bulk of the added capacitive elements is small compared to the bulk of the inductive elements. , in particular compared to the space saving obtained by optimizing the inductive elements of the LC assemblies in series 12, 22.

It is then chosen to distribute the resonance frequencies f11 and f21 in the frequency band of the second harmonics HB2, for example a resonance frequency f11, f21 in each half of the band HB2.

The inductive element and the capacitive element of the third section SCT3 are not replaced by a resonant LC assembly, in order to maintain a certain level of attenuation at the high frequencies.

In summary, the resonant frequencies f22, f12 are first chosen for each LC series assembly so as to be distributed in different harmonic frequency bands of the fundamental frequency band, and so that the series LC connections replacing the capacitive elements of the virtual matching network having the smallest capacitive values, have the largest resonance frequencies.

Then, the resonance frequencies are chosen for each LC assembly in parallel 11, 21 so as to be distributed, with the resonance frequencies of the LC assemblies in series 12, 22, in different frequency bands of harmonics of the frequency band fundamental.

Graph 52 shows that the maximum loss in the fundamental band FB is of the order of 1.6dB.

Graphs 53 and 54 show that the real part of the impedance is contained at approximately 10% around for example 4.5 ohms, and that the imaginary part is also well contained at approximately 15 pF (pico farads).

Thus, it has been described an adaptation and filtering technique advantageous both in terms of size and performance, with very inexpensive passive component assemblies.

In summary, the matching is performed by a minimum quality factor three section low pass filter to provide the impedance transformation of the real part.

The imaginary part of the optimum impedance of the power amplifier is achieved by the inductive element of the DC power supply stage.

A capacitive coupling element is added before the antenna to block the DC voltage. Its value is chosen large enough to have little impact on the transformation of the impedance.

The rejection (filtering) of the harmonics is ensured by the replacement of the inductive elements in series by LC connections in parallel, and by the replacement of the “shunt” capacitor by LC connections in series. The equivalent reactance of the LC assemblies is kept equal to the reactance of the element which they replace respectively, in the fundamental frequency band.

The resonance frequencies of the LC assemblies are advantageously chosen as follows:

- The first LC connection in parallel, on the output side of the power amplifier, blocks the upper part of the second harmonic band;

- The first LC parallel assembly blocks the lower part of the second harmonic band;

- The first LC "shunt" series circuit, on the output side of the power amplifier, removes frequencies in the third harmonic band;

- The second LC assembly in "shunt" series, evacuates the frequencies in the bands of the fourth and fifth harmonics.

- The inductive element and the capacitive element of the last section, on the antenna side, are not replaced by a resonant circuit in order to provide attenuation for higher order harmonics, i.e. say greater than five.

Claims (14)

Integrated circuit comprising a power amplifier (PA) intended to supply a signal in a fundamental frequency band, an antenna (ANT), and a matching and filtering network (MFN) comprising:
- a first section (SCT1) between an output node of the power amplifier (PA) and a first intermediate node (N1),
- a second section (SCT2) between the first intermediate node (N1) and a second intermediate node (N2), and
- a third section (SCT3) between the second intermediate node (N2) and an antenna input node (ANT),
the three sections comprising inductive-capacitive “LC” assemblies configured to present an impedance adapted to the output of the power amplifier (PA) in the fundamental frequency band,
wherein the LC assemblies of the first section (SCT1) and of the second section (SCT2) are further configured to have resonant frequencies respectively adapted to attenuate the harmonic frequency bands of the fundamental frequency band. An integrated circuit according to claim 1, wherein the LC assemblies of the first section (SCT1) are configured to have resonance frequencies lower than the resonance frequencies of the corresponding LC assemblies of the second section (SCT2). Integrated circuit according to one of claims 1 or 2, wherein the first section (SCT1) comprises a parallel LC circuit (11) coupled between the output node of the power amplifier (PA) and the first intermediate node ( N1) and a series LC circuit (12) coupled between the first intermediate node (N1) and a ground node (GND), and the second section (SCT2) has a parallel LC circuit (21) coupled between the first node intermediate (N1) and the second intermediate node (N2) and a series LC circuit (22) coupled between the second intermediate node (N2) and the ground node (GND). An integrated circuit according to claim 3, wherein each series LC circuit (12, 22) is configured to have an equivalent impedance corresponding to an impedance of a capacitive element (C ₀ ) suitable for said impedance matching in the band. fundamental frequencies, and in which the resonant frequencies (f12, f22) of each LC series assembly (12, 22) are chosen so as to be distributed in different harmonic frequency bands (HB5, HB4, HB3) of the fundamental frequency band, and so that the series LC assemblies (12, 22) having equivalent impedances corresponding to the impedances of the capacitive elements (C ₀ ) having the smallest capacitive values, have the largest resonance frequencies (f12, f22). Integrated circuit according to Claim 4, in which the resonant frequencies (f11, f21) of each parallel LC assemblies (11, 21) are chosen so as to be distributed, with the resonance frequencies (f12, f22) of the LC assemblies in series (12, 22), in different harmonic frequency bands (HB2) of the fundamental frequency band. Integrated circuit according to one of claims 3 to 5, in which the LC assemblies of the first section (SCT1) and of the second section (SCT2) are configured according to at least one of the following criteria:
- the parallel LC assembly (11) of the first section (SCT1) is configured to have a resonant frequency (f11) in one half of the frequency band of the second harmonics (HB2);
- the parallel LC assembly (21) of the second section (SCT2) is configured to have a resonant frequency (f21) in the other half of the frequency band of the second harmonics (HB2);
- the series LC assembly (12) of the first section (SCT1) is configured to have a resonant frequency (f12) in the frequency band of third harmonics (HB3);
- the LC assembly in series (22) of the second section (SCT2) is configured to have a resonant frequency (f22) either between the frequency band of the fourth harmonics (HB4) and the frequency band of the fifth harmonics (HB5) , or in a common portion of the frequency band of the fourth harmonics (HB4) and of the frequency band of the fifth harmonics (HB5). Integrated circuit according to one of the preceding claims, in which the third section (SCT3) comprises an LC assembly comprising an inductive element coupled between the second intermediate node (N2) and the input node of the antenna (ANT), and a capacitive element coupled between the antenna input node (ANT) and a ground node (GND), the LC assembly of the third section (SCT3) being configured to have a minimum quality factor. Method of impedance matching and filtering between an output of a power amplifier (PA) supplying a signal in a fundamental frequency band and an antenna (ANT), comprising dimensioning of a matching network (MN0 ) virtual comprising:
- a first section (SCT01) between an output node of the power amplifier (PA) and a first intermediate node (N1),
- a second section (SCT02) between the first intermediate node (N1) and a second intermediate node (N2), and
- a third section (SCT03) between the second intermediate node (N2) and an antenna input node (ANT),
the three sections each comprising an inductive element and a capacitive element, the dimensioning being carried out so as to present an impedance suited to the output of the power amplifier (PA) in the fundamental frequency band,
wherein the method comprises making an actual matching and filtering (MFN) network comprising a replacement of each inductive element and each capacitive element of the first section (SCT01) and of the second section (SCT02) of the network virtual adaptation (MN0), by respective inductive-capacitive "LC" assemblies (11, 12, 21, 22), configured to present an equivalent impedance adapted to the output of the power amplifier (PA) in the band of fundamental frequencies, and in addition to have resonant frequencies (f11, f12, f21, f22) respectively adapted to attenuate the harmonic frequency bands of the fundamental frequency band. Method according to Claim 8, in which the resonance frequencies of the LC assemblies (11, 12) of the first section (SCT1) of the real matching and filtering network (MFN) are chosen lower than the resonance frequencies of the LC assemblies ( 21, 22) correspondents of the second section (SCT2) of the real adaptation and filtering network (MFN). Method according to one of Claims 8 or 9, in which the first section (SCT01) of the virtual adaptation network (MN0) comprises an inductive element coupled between the output node of the power amplifier (PA) and the first intermediate node (N1) and a capacitive element coupled between the first intermediate node (N) and a ground node (GND), and the second section (SCT02) of the virtual adaptation network (MN0) has an inductive element coupled between the first intermediate node (N1) and the second intermediate node (N2) and a capacitive element coupled between the second intermediate node (N2) and the ground node (GND),
and wherein said realization of the actual matching and filtering network (MFN) comprises a replacement of each inductive element with an LC connection in parallel (11, 21), and a replacement of each capacitive element with an LC connection in series ( 21, 22). Method according to Claim 10, in which the resonant frequencies (f12, f22) are firstly chosen for each LC assembly in series so as to be distributed in different harmonic frequency bands (HB5, HB4, HB3) of the fundamental frequency band, and so that the series LC assemblies (12, 22) replacing the capacitive elements of the virtual matching network (MN0) having the smallest capacitive values, have the largest resonance frequencies. Method according to Claim 11, in which the resonant frequencies (f11, f21) are then chosen for each LC connection in parallel (11, 21) so as to be distributed, with the resonance frequencies (f12, f22) of the LC connections in series (12, 22), in different harmonic frequency bands (HB2) of the fundamental frequency band. Method according to one of Claims 10 to 12, in which the realization of the real adaptation and filtering network (MFN) is carried out according to at least one of the following criteria:
- the parallel LC assembly (11) of the first section (SCT1) has a resonant frequency (fr11) in one half of the frequency band of the second harmonics (HB2);
- the parallel LC assembly (21) of the second section (SCT2) has a resonant frequency (fr21) in the other half of the frequency band of the second harmonics (HB2);
- the LC assembly in series (12) of the first section (SCT1) has a resonant frequency (fr12) in the frequency band of the third harmonics (HB3);
- the LC assembly in series (22) of the second section (SCT2) has a resonant frequency (fr22) either between the frequency band of the fourth harmonics (HB4) and the frequency band of the fifth harmonics (HB5), or in a common portion of the fourth harmonic frequency band (HB4) and the fifth harmonic frequency band (HB5). Method according to one of Claims 8 to 13, in which the realization of the real adaptation and filtering network (MFN) comprises a reproduction of the third section (SCT3) of the virtual adaptation network (MN0) comprising an inductive element coupled between the second intermediate node (N2) and the antenna input node (ANT), and a capacitive element coupled between the antenna input node (ANT) and a ground node (GND), the LC assembly of the third section (SCT3) being dimensioned to have a minimum quality factor.